This invention pertains to removal of phenol, formaldehyde, and melamine from waste water wash streams.
In the manufacture of phenol-formaldehyde resins necessary equipment wash produces a waste wash water stream containing too much phenol and formaldehyde to be discharged into the sewers or into natural run-off waters such as rivers or lakes.
In the same way in the manufacture of melamine-formaldehyde resins equipment wash produces waste water wash streams containing too much formaldehyde and melamine to be dumped into our natural drainage. In many states governmental regulations require removal of phenol and formaldehyde and it would be desirable to remove melamine to reduce the biological oxygen demand in our natural run-off streams.
Since, in some instances, phenol-formaldehyde and melamine-formaldehyde resins are produced in the same plant a process for removing phenol and formaldehyde from the waste water wash stream from the phenol-formaldehyde resin plant and melamine and formaldehyde from the waste wash water from the melamine formaldehyde plant would be desirable. Accordingly it is an objective of this invention to most economically remove phenol, formaldehyde and melamine from waste water wash streams from both phenol-formaldehyde and melamine-formaldehyde plastic manufacturing plants. A further objective is to be able to treat either stream or both streams in a most economical manner. Normally only a few thousand gallons per week from a plant must be treated and in such small volumes carbon adsorbtion may be used to remove harmful organics and allow for easy incineration or landfill disposal of the organic loaded carbon. Melamine and phenol are adsorbed on carbon beds but formaldehyde itself is not. However, we found that formaldehyde may be reacted with ammonia to form hexamethylene tetramine or urotropine which does adsorb on carbon. The novel approach lies in securing full loading of the carbon to remove phenol; in removing or modifying interfering reactants to secure a sufficient urotropination reaction; in securing low cost precipitation removal of much of the melamine-formaldehyde in the melamine-formaldehyde stream; in securing a precipitation removal of much of phenol-formaldehyde in that stream; in taking a recycle stream from exit the phenol removal carbon beds to use as a primary equipment wash and completing equipment wash with a small fresh water stream; using a recycle wash stream in the melamine-phenol equipment wash with a small clean water wash in order to allow a precipitation removal of much of the melamine-phenol; and using a formaldehyde urotropination reaction catalyzed by carbon ahead of final carbon clean up beds; these carbon clean up beds then causing completion of the urotropination reaction and removing dissolved melamine and the urotropine formed by reacting the formaldehyde with ammonia. Thus the final process could be considered a treatment of one stream to remove some phenol and formaldehyde; the treatment of the other to remove formaldehyde and melamine in combination and a final urotropination of either or both streams to change formaldehyde to a state to be removed by carbon adsorption.
We have considered the following patents that are relevant to parts of our process:
______________________________________ No. 1,866,417 Mackert 1932 No. 3,855,123 Strudgeon 1974 No. 3,869,387 Vargia 1975 No. 4,113,615 Gorboty 1978 No. 4,216,088 Juferov 1980 No. 79,109,249 Japan, Kokai, Tokkyo Koho 1979 No. 7,890,178 Japan, Kokai 1978 ______________________________________
We have described the novel approach developed to economically remove formaldehyde and phenol from one waste water wash stream and to remove melamine and formaldehyde from a second waste water wash stream. In the following paragraphs we will describe pertinent research work: Consider first the wash stream from washing phenol-formaldehyde plastic manufacturing equipment. Commonly a caustic type cleaner is mixed with water to wash the equipment. We found that we could adjust the pH of this stream to about 8 and then hold up with stirring to produce a precipitate from reacting the phenol and formaldehyde. After filtering this solution to remove the precipitate the liquid would contain approx. 1000-1500 ppm phenol and up to 100 ppm formaldehyde. After removing phenol in carbon beds we could recycle the stream containing formaldehyde to do the bulk of the equipment wash and follow with a fresh water clean up wash to complete the washing. This recycle would increase formaldehyde concentration to over 1000 ppm. This allowed decreasing the flow from this system to the next step which was reaction of the formaldehyde with ammonia; a simple inexpensive reaction that could be done in relatively small equipment when we found that carbon would catalyze the reaction. Theoretical considerations indicate that the reaction of formaldehyde with ammonia would most nearly approach complete formaldehyde removal with
1. low temperature PA0 2. high ammonia ion concentration PA0 3. high pH or low hydrogen ion concentration PA0 A. pretreatment of the phenol-formaldehyde wash stream by adjusting pH to less than 9; holding up with stirring to partially precipitate the phenol and formaldehyde; separating the precipitate; carbon bed treating the separation effluent to remove phenol; recycling a portion of the phenol free effluent for equipment primary wash; and feeding a portion to a urotropination step where pH is adjusted to above 10 in the presence of a large excess of ammonia to react with formaldehyde to form urotropine PA0 B. pretreatment of the melamine-formaldehyde stream at pH of about pH 4.5 with hold up for about 48 hours to precipitate melamine and formaldehyde reaction products or hold up for about 2 hours above 50.degree. C. to cause a similar precipitation; filtration and partial recycle of the filtrate with the remainder being sent to a urotropination step to react formaldehyde with an excess of ammonia at a pH above 10 PA0 C. treatment of either or both streams from the pre-treatment steps by adding sufficient ammonia, which may be in the form of ammonium chloride, to be equal to about five moles of ammonia per mole of formaldehyde; adjusting pH to above 10 with sodium hydroxide; adding powdered carbon; holding up the described reactants in a stirred vessel for about 10 hours and finally feeding through carbon beds PA0 D. arranging carbon beds in series to allow complete loading of the first replaceable carbon bed to minimize use of carbon.
Theoretical calculations would indicate the following relationships when treating a solution containing 900 parts per million formaldehyde at 25.degree. C. with various ammonia concentrations and various pH values:
______________________________________ Ammonia Ion Final unreacted (moles/liter) pH Formaldehyde (ppm) ______________________________________ 30 8 0.3 0.3 10 0.3 0.03 11 0.3 ______________________________________
Such theoretical calculations indicate the final or equilibrium point of the reaction but do not indicate speed of the reaction. A catalyst, by definition, speeds up a reaction but does not change the final equilibrium. In this case the final equilibrium of 0.3 ppm would be more than sufficient removal of formaldehyde to allow discharge to the normal run off stream.
Activated carbon was tested for catalytic activity with results as follows:
______________________________________ Reaction Final Initial Ammonia/ Time @ Gms C HCHO HCHO Conc. HCHO 25 C Added/ Conc (ppm) pH Ratio (hrs) Liter (ppm) ______________________________________ 500 11 5:1 16 0 8 500 11 5:1 16 0 13 500 11 5:1 16 0 17 500 11 5:1 16 50 3 500 11 5:1 16 50 5 500 11 5:1 16 50 3 ______________________________________
From the final column showing residual formaldehyde we see that with 50 grams of carbon per liter in three samples and no carbon in the other three that the average residual formaldehyde without carbon was about 13 ppm while with carbon the formaldehyde residual was about 4 ppm. all conditions other than carbon were the same in all six samples.
Activated carbon used in the urotropination experiments was re-used several times with little loss of catalytic activity as evidenced by continued low residual formaldehyde.
In another test a solution containing 500 ppm formaldehyde was divided into two equal parts; each part treated with 1:1 moles of ammonia per mole of formaldehyde; each adjusted to pH 10 with sodium hydroxide and 15 grams of carbon per liter was added to one part. Free formaldehyde in ppm in each sample diminished as follows:
______________________________________ Time Part 1 Part 2 (hrs) (no carbon) (with carbon) ______________________________________ 0 500 500 ppm 1/2 200 125 ppm 1 175 100 ppm 2 175 60 ppm 3 175 55 ppm 4 175 50 ppm ______________________________________
Note that in four hours with the conditions outlined that the reaction had leveled out with about 175 ppm residual formaldehyde with no added carbon; with only 15 grams per liter of carbon the residual formaldehyde was reduced to 50 ppm with the reaction continuing.
These results further confirmed catalytic activity of carbon and also, when compared with previous results, desireability of pH 11 and higher mole ratio of ammonia to formaldehyde.
Testing of fresh field sample containing melamine and formaldehyde indicated interference and variable results when treated at pH 11 with 5:1 moles of ammonia per mole of formaldehyde so that direct reaction to form urotropin from the formaldehyde was not feasible. However samples held in barrels more than one month and filtered indicated total dissolved solids of about 10 mg./ml. consistently and could be treated to form a urotropine as described above. Now in order to speed up the "aging process" we found that we could adjust the pH to 4.5 and heat above 50.degree. C. for more than 2 hours and filter. Again we found the filtrate to contain approx. 10 mg/ml of total dissolved solids; further the filtrate could then be successfully treated to tie up the formaldehyde as urotropine.
Further we found that fresh samples could be adjusted to pH 4.5 and held at room temperature for 2 days, filtered and successfully treated. Experimentation indicated pH of 4.5 to be the optimum pH to secure greatest precipitation of the interfering melamine-formaldehyde.
Using the treatment described we found that approx 70 liters of urotropination reactor effluent could be fed through 1200 grams of carbon in a first 3" column and through 930 grams of carbon in a second 3" column at the rate of approx. 95 ml/min before formaldehyde in the effluent exceeded 3 ppm. The feed to the column contained 1% melamine, 230 ppm urotropine, 23 ppm formaldehyde, approx. 3000 ppm sodium hydroxide and 3500 ppm ammonium chloride. Results showing leakage of formaldehyde and urotropine are shown in Table 1 below
TABLE 1 __________________________________________________________________________ Fixed Bed Carbon Adsorption of Urotropin 1% Melamine Flow Cum. Column I Column II Cum. Time Rate Flow Urotropine HCHO Urotropine HCHO TDS hrs ml/mn ltr ppm ppm ppm ppm mg/g __________________________________________________________________________ 0 97.0 0 -- -- -- -- 1 95.5 5.8 3.9 0 2.9 0.0 4.2 2 96.5 11.6 -- -- 4.7 -- 4.3 3 96.0 17.3 3.9 8.4 4.8 0.0 4.3 4 94.0 23.0 -- -- 8.6 -- 4.3 5 84.0 28.0 39 7.2 3.9 1.2 4.3 6 88.0 33.3 -- -- 8.6 -- 4.3 7 96.0 39.1 125 13.2 8.2 0.0 4.3 8 94.5 44.7 -- -- 11. -- 4.3 9 89.0 50.1 159 9.6 4.0 2.4 4.3 10 92.0 55.6 -- -- 11. -- 4.3 11 92.0 61.1 184 8.4 30. 1.2 4.3 12 90.0 66.5 -- -- 76. -- 4.5 13 91.0 72.0 229 11.0 125 3.6 4.8 14 91.5 77.5 -- -- 124 -- 5.3 15 88.0 82.7 187 9.6 154 7.2 5.7 16 90.0 88.1 -- -- 136 -- 6.1 17 87.5 93.4 219 9.6 222 7.2 6.4 18 91.0 98.8 -- -- -- -- 6.7 19 89.5 104.2 210 9.6 202 14 6.8 20 86.0 109.4 -- -- -- -- 7.0 21 91.0 114.8 -- -- 222 16 -- 22 90.0 120.2 -- -- -- -- -- 23 89.5 125.6 -- -- 222 12 -- 24 88.0 130.9 -- -- -- -- -- __________________________________________________________________________
Note that the inlet feed contained 23 ppm formaldehyde along with excess ammonia and sodium hydroxide or proper chemical conditions for further reaction of formaldehyde and ammonia.
Consideration of these results show that:
a. the total dissolved solids of 4.3 mg/gm exactly equals the sodium chloride which would be formed from the ammonium chloride and sodium hydroxide in the feed; the ammonium hydroxide formed would volatilize during evaporation in the analytical process to determine total dissolved solids; thus we see that all the melamine was removed until about seventy-two liters of the stream was processed; because melamine, if present, would add to the total dissolved solids.
b. formaldehyde itself does not adsorb on carbon so that removal of formaldehyde proceeded through reaction to urotropine and adsorption of the urotropine on the carbon thus driving the urotropination to completion or by catalytic activity of additional carbon in the beds or by combination of these mechanisms.
Thus carbon bed removal of urotropine removes melamine, urotropine, and completes the urotropination reaction to remove formaldehyde and remove the additional urotropine formed.